Senior Computer Vision Engineer

At Digital Place Vision, we are a team of Visionaries on a mission to light up the retail world with intelligent AI. We are seeking an exceptional Senior Computer Vision Engineer to help stores do planogram compliance.

In this role, you will lead the development of end-to-end computer vision algorithm systems that enable retail stores to accurately locate and recognize every SKU (around 5000+ SKUs) on the shelf. From building robust visual recognition pipelines for complex, real-world retail environments to leveraging AI for automated product detection, classification, and localization, you will develop intelligent systems that assess planogram compliance and generate actionable recommendations to improve store execution and operational efficiency.

Responsibilities:

● Lead the design, development, and deployment of end-to-end computer vision systems for retail shelf analysis and planogram compliance.

● Develop and optimize algorithms for product detection, SKU recognition, localization,and shelf segmentation in complex retail environments.

● Build scalable AI pipelines that accurately identify products and evaluate planogram compliance from images and video captured in stores.

● Design and implement machine learning and deep learning models for SKUs identification and detection.

● Improve model accuracy, inference speed, and robustness across varying lighting conditions, camera angles, shelf layouts, and product packaging.

● Develop data collection, annotation, augmentation, and model evaluation strategies to continuously improve AI performance.

● Analyze model performance using quantitative metrics and conduct error analysis to identify opportunities for improvement.

● Research and evaluate the latest advancements in computer vision, and deep learning, and translate them into production-ready solutions.

● Optimize models for edge devices (Android device) and cloud-based inference to ensure scalable and efficient deployment.

● Establish engineering best practices, including code quality, testing, documentation, and MLOps workflows.

● Drive technical innovation and contribute to the long-term vision of the company's retail AI platform.

● Field Testing & Iteration: Conduct extensive testing in real retail warehouse environments. Analyze edge cases, continuously iterate, and improve the recall and accuracy rates of the vision system.

Requirements:

● Education: Bachelor’s degree or above in Computer Science, Artificial Intelligence, or a related field.

● Core Programming Skills: Solid programming foundation with proficiency in C++ and Python. Familiarity with Linux development environments and strong software engineering practices.

● Traditional Vision & Camera Tech: Strong mastery of OpenCV and a deep understanding of fundamental to advanced image processing algorithms. Nice to have: proven hands-on experience with camera calibration, multi-camera synchronization, and handling complex lighting/motion blur.

● MachineLearning/Deep Learning (ML/DL) Expertise:

o Proficiency in mainstream deep learning frameworks (e.g., PyTorch or TensorFlow).

o Ability to design custom vision models using PyTorch or TensorFlow if needed.

o Understanding of best practices on model tuning to get best results.

o Ability to do benchmarking based on the real cases, so the benchmarking result should align with the real case that happens in the field.

o Ability to analyze image dataset such as using FiftyOne and create custom modules or functions to enhance the analysis.

o Understanding how to inspect and analyze models using tools such as Grad-CAM, MLFlow, Tensorboard, Wandb, etc (Interpretable AI).

● Edge Deployment Experience: Familiarity with AI model on edge deployment and inference acceleration tools such as TensorRT, ONNX Runtime, NCNN, or CUDA. Knowing how to extract maximum performance from compute-constrained devices is crucial. Especially in Android Device, expected to know how to fully utilize GPU on Android device for batch inferences.

● MLOps & Model Lifecycle Management: Experience with MLOps platforms such as Weights & Biases (W&B), MLflow, TensorBoard, or similar tools for experiment tracking, model versioning, and reproducible training workflows. Familiarity with continuous model evaluation, including confidence threshold tuning, uncertainty estimation, new SKU (novelty/OOD) detection, model drift monitoring, and automated performance benchmarking is highly desirable.

● Monitoring & Observability: Experience implementing production monitoring and logging for AI systems using tools such as Grafana, Prometheus, ELK Stack, OpenTelemetry, or similar observability platforms. Ability to monitor inferencelatency, model accuracy, system health, and business KPIs, while proactively identifying performance degradation and production issues.

● Cloud Deployment: Experience deploying and managing machine learning workloads on cloud platforms such as Google Cloud Platform (GCP), AWS,or Microsoft Azure. Familiarity with cloud-native services, containerization (Docker), orchestration (Kubernetes), CI/CD pipelines, scalable inference services, and MLOps infrastructure for production AI systems is preferred.

Nice to have:

● Robotics Middleware: Familiarity with ROS / ROS2, understanding node communication, and image transport mechanisms (e.g., image_transport, cv_bridge).

● Bonus Skills: Experience using professional barcode decoding libraries (e.g., ZBar, ZXing, Halcon). Background in developing CV systems for the retail inventory or warehousing/logistics industries.